This invention relates generally to spread spectrum communication systems and, more particularly, to a method and apparatus for despreading Offset-QPSK (OQPSK) spread signals in a DS-CDMA receiver.
Traditionally, mobile radio communication systems have been optimised mainly towards speech services and just marginally towards data communication services. However, as an increased number of subscribers in the existing networks as well as requests for new kinds of services has emerged, capacity problems in the existing networks are foreseen. To meet the new flexibility and capacity requirements DS-CDMA (Direct Sequence Code Division Multiple Access) based technology has turned out to be a promising candidate for the choice of multiple access method adapted to the air interface of future mobile radio communication systems.
Compared to traditional systems where the users are separated by use of narrow or seminarrow frequency band combined with or without time division, in CDMA based systems each user is assigned a different pseudo-noise spreading sequence. This gives rise to a substantial increase in bandwidth of the information-bearing signal. Spread spectrum systems generally fall into one of two categories: frequency hopping (FH) or direct sequence (DS). This invention relates to DS-CDMA. Direct sequence is, in essence, multiplication of a conventional communication waveform by a pseudo-noise, real or complex, sequence in the transmitter.
Thus DS-CDMA systems use real or complex (polyphase) sequences as means to spread the bandwidth of a transmitted signal in order to achieve simultaneous operation of multiple users in the same frequency band. As is common in the art the complex spreading sequences (pseudo-noise) occupying the same spectrum for the different users are chosen to have certain correlation properties in order to interfere with each other as little as possible. In each receiver adapted to receive the transmitted signal in question the inverse operation of spreading the transmitted signal spectrum, called despreading, is performed in order to recover the original data signal and in the same time suppress the interference from the other users or, more generally, the other sources.
The despreading operation is performed prior to data demodulation and decoding and it is also the basic operation within the multipath delay search processor (searcher), which is an important part of a so-called RAKE receiver. The searcher is used to estimate the channel impulse response, to identify paths within a delay profile and to keep track of changing propagation conditions. A RAKE receiver should be able to capture most of the received signal energy by allocating a number of parallel demodulators to the selected strongest components of the received multipath signal. The allocation and time synchronisation of the demodulators are performed on the basis of the estimated channel response.
If the spreading sequence is real and binary (xc2x11 element) then the spreading is called BPSK spreading (BPSK=Binary Phase Shift Keying), while when the spreading sequence is complex, i.e. consists of real and imaginary components which are both considered to be binary sequences, it is called QPSK spreading (QPSK=Quadrature Phase Shift Keying). Usually QPSK spreading is performed by multiplying the data with different real and imaginary binary sequences as described in the article xe2x80x9cA Coherent Dual-Channel QPSK Modulation for CDMA Systemsxe2x80x9d, by S. R. Kim et al, Proc. of VTC""96, Atlanta, pp 1848-1852. April 1996. Pulse shape filtering is usually performed in each of the quadrature branches of the system in order to adapt the spread signal to the transmission channel.
It is to be noted that QPSK spreading could be applied to either BPSK or QPSK data modulation format. Examples of both are given in the article mentioned above. For both data modulation formats the QPSK despreader is the same, i.e. multiplication of quadrature input samples with the complex conjugated spreading sequence and integration across a data symbol period.
This invention relates to a receiver adapted to receive signals with arbitrary data modulation formats spread by using Offset-QPSK (OQPSK) spreading. The OQPSK spreading differs from QPSK spreading by a half chip period delay in the imaginary (Q) branch of the spreader, after multiplication of the data symbol with the segment of complex spreading sequence. OQPSK spreading is used in the up-link of the so called IS-95 systems and is also discussed for third generation mobile systems.
A despreader for BPSK data modulation with Offset-QPSK spreading is discussed in D. M. Grieco, xe2x80x9cThe Application of Charge-Coupled Devices to Spread-Spectrum Systems,xe2x80x9d IEEE Transactions on Communications, Vol. 28, No. 9 (Chapter IIIC, pp 1699, FIG. 7). Another reference discussing the same matter is D-W Lee et al, xe2x80x9cDevelopment of the Base Station Transceiver Subsystem in the CDMA Mobile Systemxe2x80x9d, ETRI Journal, Vol. 19, No. 3, pp 116-140, October 1997.
FIG. 1 illustrates a prior art despreader according to the references mentioned above. A received signal is downconverted into its corresponding baseband representation and divided into inphase yI and quadrature yQ signal components. The signal components are further downsampled through downsampling means A1 and A2, respectively, to provide two complex samples per chip period, Tc, i.e. the duration of a complex PN sequence symbol. The received signal components are multiplied by multipliers A3, A4, A5, A6 with the real and imaginary parts dI(n) and dQ(n), respectively of the corresponding complex PN sequence symbol. Note that the signal components multiplied with the real PN sequence symbol component dI(n) are delayed half a chip period, in delay circuits A7 and A8, respectively, before the actual multiplications are performed. This is done in order to align the signal components. The resulting multiplied signals are combined, in combination circuits A9 and A10, respectively, downsampled with a factor 2, in downsampling means A11 and A12, respectively, and fed to the inputs of two summation circuits A13 and A14, respectively, which are performing a correlation operation. If the PN sequence of the received signal and the local replica generated by the receiver are synchronised, as will be described further on, the correlator provides a constructively combined signal which can be used for data demodulation and detection.
Assuming perfect synchronisation, and that the data signal and the complex PN sequence can be represented as
d(n)=dI(n)+jdQ(n) and s(k)=sI(k)+jsQ(k)
respectively, it can be derived that the output from the correlator can be represented as:                               z          ⁡                      (            n            )                          =                  xe2x80x83                ⁢                              2            ⁢                          L              ·                              {                                                                            d                      r                                        ⁡                                          (                      n                      )                                                        +                                                            jp                      ⁡                                              [                                                                              T                            c                                                    /                          2                                                ]                                                              ·                                                                  d                        Q                                            ⁡                                              (                        n                        )                                                                                            }                                              +                      2            ⁢                                          d                Q                            ⁡                              (                n                )                                      ⁢                          ∑                              [                                                                            s                      i                                        ⁡                                          (                      k                      )                                                        ⁢                                                            s                      Q                                        ⁡                                          (                      k                      )                                                                      ]                                              +                                                  xe2x80x83                ⁢                  j          ⁢                      xe2x80x83                    ⁢                                    ∑              L                        ⁢                          [                                                                                          W                      ⁡                                              (                        k                        )                                                              Q                                    ⁢                                                            s                      i                                        ⁡                                          (                      k                      )                                                                      -                                                                            W                      ⁡                                              (                        k                        )                                                              t                                    ⁢                                                            s                      Q                                        ⁡                                          (                      k                      )                                                                                  ]                                          
each integration (xcexa3) is made from k=nL to k=nL+Lxe2x88x921, where L is equal to the data symbol duration in chip periods. From this equation it can be seen that the inphase and quadrature components of the data signal d(n) are weighted differently. The quadrature component is weighted by a factor determined by the impulse response of the used pulse shaping filter. This constitutes no problem for the BPSK modulated signals where the quadrature phase signal usually carries no information, i.e. dQ(n)=0. However, for signals utilising quadrature based modulation schemes, e.g. QPSK modulated signals, such weighting will make data demodulation difficult since the real and the imaginary parts of each data modulation symbol can not be demodulated with the same quality. It can also be seen from the equation above that the self-interference due to crosscorrelation between the real and imaginary PN sequences increase when quadrature data modulation is applied. Furthermore, the self-interference due to crosscorrelation between the real and imaginary PN sequences as well as the distortion due to interfering chip waveforms originating from adjacent chip intervals will be unevenly distributed between the real and imaginary parts of z(n). All these factors implies a reduced performance of the overall system if quadrature modulated signals are utilised with the despreader scheme specified.
Thus, summarizing, the existing OQPSK despreader known in the prior art has two major drawbacks when it is applied to the OQPSK spread-spectrum signals obtained for data modulation formats other than BPSK, i.e.:
a) The data bits contained in the real and imaginary parts of each data modulation symbol can not be demodulated with the same quality since the amplitudes after despreading are different in the real and imaginary part of the data modulation symbol due to the pulse shaping filter operation;
b) The self-interference due to the crosscorrelation of real and imaginary PN sequences is increased compared to the case when BPSK/OQPSK modulation/spreading is applied.
An object of the invention is to provide an Offset-QPSK despreader scheme which can function independently of the data modulation format. Thus, the despreader in the receiver does not need to be changed if a change in data modulation format is done in the transmitting system, for instance in new systems developed in the future.
Another objective of the invention is to provide a computational efficient OQPSK despreader scheme tractable from an implementation point of view.
The object of the invention is thus to provide a despreader scheme valid for OQPSK spreading and QPSK and BPSK or any other complex data modulation format.
The technical field for the invention is to provide a method and a device for despreading incoming signals in a receiver for DS-CDMA systems (DS-CDMA=Direct Sequence Code Division Multiple Access) using Offset-QPSK spreading as the transmitting spreading format combined with arbitrary data modulation formats, such as BPSK and QPSK data modulation formats. The receiver has a unique PN sequence signal divided into a real and an imaginary component. The input base-band signal is divided into an inphase and a quadrature component, as common in the art. The invention is characterised in that complex correlations are performed between an OQPSK data signal and corresponding segments of a locally generated replica of the PN sequence used to spread the data signal. The invention is further characterised in that the inphase and the quadrature components of the input signal are downsampled to provide either two samples per chip period or one complex sample per chip period in which later case the down-sampling instants are delayed by a quarter of the chip period. Each downsampled complex signals is correlated with the corresponding segments of the complex PN sequence signal of the receiver. The correlation values obtained for odd and even complex samples of the downsampled input signal are summed, in the case when two samples per chip are provided after downsampling.
Each of the samples of the inphase and the quadrature components of the input signal, or a timeshifted version of them, could be multiplied with each of the real and an imaginary component of the PN spreading sequence of the receiver. The output of the multiplied inphase components version with the real PN spreading sequences and the output of the multiplied quadrature components with the imaginary PN spreading sequence are added to each other. The output of the multiplied quadrature components version with the imaginary PN spreading sequence and the output of the multiplied inphase components version with the imaginary PN spreading sequences are subtracted one from the other.
The invention was developed in order to provide a receiver adapted for systems using OQPSK spreading format. The inventive despreader has a good performance on AWGN and Rayleigh fading channels. It is valid for BPSK, QPSK or any other complex data modulation formats, which makes it important for future mobile communication systems.